review-code

To review a pull request for code health, you can run the review-code skill on your latest PR. It analyzes diffs to identify design, correctness, and security issues, then tags each finding by severity so high-priority problems are addressed first before merging.

Automated PR analysis evaluates code changes for design, correctness, and security flaws. This review process grounds each finding directly in the codebase and assigns severity tags, ensuring that the most critical issues are surfaced and addressed before lower-priority suggestions.

No, performing thorough code review and PR analysis requires access to the codebase and knowledge of the project's security model. Grounding findings in the actual code and assessing correctness accurately depends on having full contextual access to the repository structure.

Yes, automated PR review explicitly checks code changes for security vulnerabilities. By analyzing the diff against the project's known security model, it identifies potential security flaws and tags them as priority issues to ensure they are resolved before the code is merged.

The best way to ensure code correctness before merging is to run an automated review on the PR diff. This process evaluates the changes for design flaws and correctness issues, tagging findings by priority so you can systematically resolve high-severity problems first.

Automated code health analysis for pull requests requires direct access to the codebase and an understanding of the project's security model to function accurately. Without this context, the review cannot effectively ground its findings or properly assess design and correctness issues.